On 2009-08-26 16:48 +0200, Edward Welbourne wrote: > Apologies for the empty report text; somewhere between reportbug, > emacsclient and vi, I wasn't given the opportunity to edit the report > before sending it ! My EDITOR=emacsclient didn't work, for reasons > unknown (and, apparently, specific to emacs23) so I told reportbug to > use vi; but (I now know) it somehow managed to invoke vi in a way that > left me editing a file called -c (in reportbug's working directory), > that didn't contain the package description; and, when saved, hadn't > changed the file reportbug was expecting me to edit; so I was obliged > to over-ride reportbug's unwillingness to submit the bug without an > edited body. > > When I came to report the bug in reportbug, it threw me into > emacsclient (now working, as I've exited emacs23 and fallen back on > emacs22), but I found myself editing a buffer called -dir, on a file > in the working directory of reportbug - and, alongside it, was a file > named -c that contained the text I'd prepared, using vi, for this bug: > here's what it said <quote> > > *sigh* > Additionally, I must sadly add that emacsclient doesn't seem to be working. > I *would* be editing this using emacsclient otherwise, but I'm forced to > use vi, because that at least *works* (damnit). > I restarted emacsclient in my running emacs23, to no avail.
Unfortunately, the emacs23 emacsclient is not compatible with the emacs22 server, and vice versa. However, "M-x server-start" in emacs23 should work, unless you already have a server running from another Emacs session. Previously, server-start would have silently overtaken the server socket from previous instances, but in Emacs 23 you need to call server-force-delete. Here is what I have in my .emacs to accommodate for that situation: (when (>= emacs-major-version 23) (require 'server) (unless (server-running-p) (server-start))) > Anyway, the error message I get when I try to load my mailbox says: > > Wrote /disk/home/eddy/.sys/tmp/rmail738ElS > Writing messages to /disk/home/eddy/.sys/tmp/rmail738mIt... > byte-code: Removing old name: no such file or directory, > /disk/home/eddy/.sys/tmp/rmail738mIt > > and I'm left with the babyl file loaded but not processed in rmail mode. > Trying to M-x rmail-mode in it, I get > > Wrote /disk/home/eddy/.sys/tmp/rmail738M7U > Writing messages to /disk/home/eddy/.sys/tmp/rmail738ZFb... > byte-code: Removing old name: no such file or directory, > /disk/home/eddy/.sys/tmp/rmail738ZFb Please send a lisp backtrace after "M-x toggle-debug-on-error". It might be useful to load the uncompiled rmail.el so that the backtrace is more meaningful. > </quote> (I note that there is ample space on /tmp/: > <quote src="df /tmp"> > Filesystem 1K-blocks Used Available Use% Mounted on > /dev/sda1 9614116 6778496 2347248 75% / > </quote> so the problem wasn't that it'd run out of diskspace.) Is /tmp on the same filesystem as /disk/home/eddy/.sys/tmp ? Sven -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
